U.S. District Judge Vince Chhabria granted summary judgment to Meta in a case brought by 13 authors, including Ta-Nehisi Coates, Sarah Silverman, and Pulitzer Prize winners Junot Díaz and Andrew Sean Greer, who alleged the company illegally used their books to train its Llama AI models. The judge found Meta's use was "highly transformative" under copyright law's fair use doctrine and that the authors failed to present adequate evidence of how they were harmed by Meta's actions.
Critically, the 40-page ruling repeatedly emphasized that the decision does not vindicate Meta's practices or provide legal cover for AI training on copyrighted works. "This ruling does not stand for the proposition that Meta's use of copyrighted materials to train its language models is lawful," Judge Chhabria wrote. "It stands only for the proposition that these plaintiffs made the wrong arguments and failed to develop a record in support of the right one."
Judge Chhabria's opinion appears designed to guide future litigation by identifying "market dilution" as potentially the strongest argument against AI training. This theory suggests AI systems trained on books could generate competing works that harm the market for human-authored books.
The court outlined three ways plaintiffs might argue market harm. First, models could regurgitate copyrighted works, allowing free access to the originals. Second, unauthorized copying could harm licensing markets for AI training. Third, models could generate works similar enough in subject matter or genre to compete with the originals.
"In this case, the first two arguments fail," the Chhabria explained. "The third argument is far more promising, but the plaintiffs' presentation is so weak that it does not move the needle."
The Chhabria offered a pointed example about publishing's vulnerability. "Take, for example, biographies. If a company uses copyrighted biographies to train a model, and if the model is thus capable of generating endless amounts of biographies, the market for many of the copied biographies could be severely harmed. Perhaps not the market for Robert Caro's Master of the Senate, because that book is at the top of so many people's lists of biographies to read. But you can bet that the market for lesser-known biographies of Lyndon B. Johnson will be affected."
AI-generated books, Chhabria argued, might not affect sales of works by well-known authors. But they could, he continued, "very well prevent the next Agatha Christie from getting noticed or selling enough books to keep writing."
The court noted that "generative AI has the potential to flood the market with endless amounts of images, songs, articles, books, and more." Yet, Chhabria wrote in his order, "the plaintiffs never so much as mentioned it in their complaint. Nor did they mention it in their own summary judgment motion." He went on to suggest that this theory could prove decisive in future cases: "It seems like the plaintiffs will often win, at least where those cases have better-developed records on the market effects of the defendant's use."
Leaning on piracy—again
Meta initially attempted to license books from publishers, with the company's head of generative AI discussing spending up to $100 million on licensing deals. But when those negotiations stalled in spring 2023, CEO Mark Zuckerberg approved using unlicensed materials instead. Economics makes it plain why Meta chose the expediency of piracy over legal licensing: the company expects its generative AI products to generate $2-$3 billion in revenue in 2025 and $460 billion to $1.4 trillion over the next decade, according to court documents.
Meta downloaded the plaintiffs' books from "shadow libraries" including Library Genesis and Anna's Archive. Meta obtained at least 666 copies of books whose copyrights the plaintiffs hold, including Sarah Silverman's comic memoir The Bedwetter, Rachel Louise Snyder's nonfiction book No Visible Bruises: What We Don't Know About Domestic Violence Can Kill Us, Junot Díaz's Pulitzer Prize–winning novel The Brief Wondrous Life of Oscar Wao, and Andrew Sean Greer's Pulitzer-winning novel Less.
Chhabria explicitly contradicted the prevailing argument in Big Tech circles that any ruling against tech companies would hinder AI development. "The suggestion that adverse copyright rulings would stop this technology in its tracks is ridiculous," Judge Chhabria wrote. "These products are expected to generate billions, even trillions, of dollars for the companies that are developing them. If using copyrighted works to train the models is as necessary as the companies say, they will figure out a way to compensate copyright holders for it."
The Meta ruling comes just a day after U.S. District Judge William Alsup ruled in Bartz v. Anthropic that while training AI on copyrighted books constituted fair use, downloading pirated copies for permanent storage violated copyright law. Judge Alsup found Anthropic's training "exceedingly transformative," but ordered a trial on the company's use of over seven million pirated books.
Judge Chhabria explicitly criticized aspects of Judge Alsup's comments comparing AI training to human learning. Chhabria found Alsup’s argument that "everyone reads texts, too, then writes new texts" insufficient, noting that Alsup was "brushing aside concerns about the harm it can inflict on the market for the works it gets trained on" and calling Alsup's analogy to teaching schoolchildren "inapt."
"Using books to teach children to write is not remotely like using books to create a product that a single individual could employ to generate countless competing works with a minuscule fraction of the time and creativity it would otherwise take," Chhabria added. A Case management conference to address Meta’s actions is scheduled for July 11.
The attorneys representing the plaintiffs suing Meta expressed frustration. "The court ruled that AI companies that 'feed copyright-protected works into their models without getting permission from the copyright holders or paying for them' are generally violating the law," attorneys at Boies Schiller Flexner said in a statement. "Yet, despite the undisputed record of Meta's historically unprecedented pirating of copyrighted works, the court ruled in Meta's favor."
Meta, on the other hand, took a more positive view. "We appreciate today's decision from the Court," Meta spokesperson Thomas Richards said. "Fair use of copyright material is a vital legal framework for building this transformative technology."
Despite the narrow victories for Meta and Anthropic, the rulings leave fundamental questions about AI training and copyright law unresolved while providing a path for future litigation. Several other high-profile cases involving OpenAI, Microsoft, and other AI companies remain pending in federal courts, with plaintiffs likely to adjust their strategies based on these early rulings.
